C MRS: An Outstanding Tool for Metabolic Studies

C MRS: An Outstanding Tool for Metabolic Studies

TIAGO B. RODRIGUES, ^1,2 SEBASTIA´N CERDA´N ¹

1

Instituto de Investigaciones Biome´dicas “Alberto Sols” C.S.I.C./U.A.M., Madrid, Spain

2

Departamento de Bioquı´mica, Centro RMN e Centro de Neurocieˆncias e Biologia Celular, Faculdade de Cieˆncias e Tecnologia, Universidade de Coimbra, Coimbra, Portugal

ABSTRACT: We provide an overview of

C magnetic resonance spectroscopy methods and their applications as an established metabolic tool with an emphasis on the use of high-resolution

C magnetic resonance spectroscopy and

C isotopomer analysis. Topics addressed include general properties of the

C magnetic resonance spectroscopy spec- trum; different

C magnetic resonance spectroscopy acquisition protocols; determination of fractional

C enrichment, kinetic, or steady state measurements of metabolic flux;

C isotopomer analysis approaches;

C(

H) magnetic resonance spectroscopy methodologies;

and in vivo

C magnetic resonance spectroscopy. Some illustrative applications are described. © 2005 Wiley Periodicals, Inc. Concepts Magn Reson Part A 27A: 1–16, 2005

KEY WORDS:

C MRS; isotopic dilution; isotopomer analysis; tricarboxylic acid cycle;

glutamine cycle; hydrogen turnover

INTRODUCTION

The first ¹³ C magnetic resonance spectroscopy ( ¹³ C MRS) study of a living organism was probably re- ported in 1972 (1). Authors followed the metabolism of (1- ¹³ C) glucose by a eukaryotic cell system and concluded that the use of this pioneering technique

“could have numerous applications for in vivo meta- bolic studies.” Since then, ¹³ C MRS has developed steadily to a method routinely used in metabolic re- search with cells, perfused organs, animals, and even humans (2–5). This progress has been favored by the

ability of ¹³ C MRS to (i) perform repetitive, nonin- vasive measurements of metabolic processes as they proceed in their own intracellular environment and (ii) its capacity to measure unique physical properties not detectable by other methodologies, such as spin cou- pling patterns, isotopic shifts, or magnetic relaxation times T ₁ and T ₂ . These properties have been shown to provide valuable information on the operation in situ of specific metabolic pathways or the dynamics of some important biological assemblies, exceeding in many cases the interpretations provided by previously used radioactive, spectrophotometric, or fluorimetric approaches.

13 C MRS allows detecting magnetic resonances from ¹³ C, the only stable isotope of carbon having a magnetic moment. The natural abundance for ¹³ C is roughly 1.1% of the total carbon and its magnetogyric ratio is approximately one-fourth of that of the proton.

These two circumstances make ¹³ C MRS an insensi- tive technique (6). Despite this, natural abundance

13 C resonances from carbons of the fatty acid chains Received 14 March 2005; revised 6 June 2005; ac-

cepted 12 June 2005

Correspondence to: Dr. Sebastia´n Cerda´n; E-mail: scerdan@iib.

uam.es

Concepts in Magnetic Resonance Part A, Vol. 27A(1) 1–16 (2005) Published online in Wiley InterScience (www.interscience.wiley.

com). DOI 10.1002/cmr.a.20039

© 2005 Wiley Periodicals, Inc.

1

of triglycerides are classically observed in most tis- sues, and glycogen carbons can be detected in the liver and muscle of fed animals and man (2, 3, 7). The sensitivity problem can be improved markedly by using ¹³ C enriched substrates. The combination of ¹³ C MRS detection and substrates selectively enriched in

13 C in specific positions have made it possible to follow in vivo and in vitro the activity of a large variety of metabolic pathways in cells, animals, and humans. These include glycolysis and the pentose phosphate pathway, glycogen synthesis and degrada- tion, gluconeogenesis, the tricarboxylic acid cycle, ketogenesis, ureogenesis and the glutamate, glu- tamine, GABA cycle in brain, and others (see refer- ences [2– 4] for reviews).

The design of ¹³ C MRS experiments with selec- tively ¹³ C-enriched substrates is similar to the classi- cal radiolabeling experiments using ¹⁴ C. A relevant difference is that ¹³ C precursors are administered in substrate amounts, whereas ¹⁴ C substrates are used in tracer amounts. However, ¹³ C MRS presents impor- tant advantages over ¹⁴ C. First, the metabolism of the

13 C-labeled substrate can be followed in real time, in situ, and noninvasively (4, 5). Second, even if tissue extracts are prepared, the detection of ¹³ C in the different carbon resonances of a specific metabolite does not require separation and carbon by carbon degradation, a prerequisite in the experiments with radioactive ¹⁴ C (8). Finally, the analysis by ¹³ C MRS of homonuclear spin-coupling patterns and isotope effects allows investigation if two or more ¹³ C atoms occupy contiguous positions in the same metabolite molecule. The latter approach represents, as will be illustrated later in this article, an enormous gain in information as compared with the classical radioac- tive ¹⁴ C experiments (8). As a counterpart to these advantages, ¹³ C MRS is significantly less sensitive than other conventional metabolic techniques, such as radioactive counting, mass spectrometry, and spectro- photometric or fluorimetric methods.

The purpose of this article is to provide an intro- duction to ¹³ C MRS methods and applications with particular reference to high-resolution ¹³ C NMR and

13 C isotopomer approaches. The methodology de- scribed here refers mainly to studies performed with tissue extracts but provides an adequate framework to understand the approaches used with in situ animals and human beings. More information on the latter aspects may be found in references (9 –12) or in NMR in Biomedicine’s 2003 special issue titled “ ¹³ C NMR Studies of Cerebral Metabolism.” Other reviews cover in more detail the metabolic information de- rived from ¹³ C isotopomer analysis in the adult mam-

malian brain (13) or in primary cultures of neural cells (14, 15).

METHODOLOGY

The ¹³ C MRS Spectrum: Spin Coupling Patterns and Isotopic Shifts

13 C resonances are distributed over a large chemical shift range and experience low relaxation rates (16).

Normally, high-resolution ¹³ C MRS spectra of metab- olites depict a collection of well-resolved narrow res- onances distributed over large chemical shift range (

250 ppm), even in viscous media as those found in vivo.

Figure 1 shows typical proton-decoupled ¹³ C MRS spectra of perchloric acid extracts obtained from the brain of rats infused with (1,2- ¹³ C ₂ ) glucose [see Fig.

1(A)] or (1,2- ¹³ C ₂ ) acetate [see Fig. 1(B)] as com- pared with the natural abundance ¹³ C MRS spectrum of a rat infused with unlabeled glucose (17 ). The most important resonances detected are those from the car- bons of glutamate, glutamine, GABA, NAA, inositol, and glucose. The increased intensity of the resonances observed in the upper panels clearly indicates that the

13 C label has been incorporated in cerebral metabo- lites. Indeed, the difference between the intensity of the natural abundance signal and the total intensity of the corresponding multiplet resonance in the spectrum obtained with ¹³ C-labeled substrate reveals the net amount of ¹³ C incorporated in the corresponding car- bon. Most of the resonances observed in the upper panel of Fig. 1 depict an apparent triplet structure produced by homonuclear ¹³ C- ¹³ C coupling. These pseudotriplets are derived from the superposition of doublets originated in those metabolites containing two contiguous ¹³ C atoms in the same molecule or from singlets, corresponding to those metabolites con- taining the ¹³ C atom bonded to ¹² C neighbors.

Figure 2 and reference (13) illustrate more clearly

these aspects. If pathway A incorporates one ¹³ C atom

from the labeled substrate in position i of metabolite

M, a singlet resonance will appear at frequency

i . If

by any chance the ¹³ C atom from the substrate is

incorporated through a different mechanism (pathway

B) in position i

1 of M, a new singlet resonance will

appear, now located at a different frequency

i

1 in

the ¹³ C MRS spectrum. It becomes then possible to

determine the relative contributions of pathways A

and B to the formation of M by the relative intensities

of the ¹³ C resonance at

i to that of

i

1 . If pathways

A and B are both active, ¹³ C atoms may be incorpo-

rated simultaneously at positions i and i

1 of the

same metabolite molecule. In this case, a new inter- action appears between the adjacent magnetic mo- ments of contiguous ¹³ C nuclei. This interaction splits the original singlet resonances into doublets. This is so because the multiplicity of ¹³ C resonance coupled to nucleus X is given by the rule 2nI

1, where n is the number of X nuclei coupled to ¹³ C, and I is their angular momentum (I

1/2 for X

¹ H, ³¹ P, ¹³ C, ¹⁵ N or I

1 for X

² H). This interaction is known as homonuclear scalar coupling and is transmitted through the electrons of the ¹³ C- ¹³ C-bond.

13 C MRS Acquisition: ¹ H Decoupling and NOE Enhancement

Figure 3 illustrates the most important characteristics of ¹³ C MRS data acquisition and the resulting ¹³ C MRS spectra for the hypothetical case of a ¹³ C carbon bonded only to a vicinal proton. Spin coupling to one or more protons complicates the interpretation of the

13 C spectra and thus it is normally removed using a variety of proton decoupling techniques. All decou- pling procedures are based on the same principle, the

broadband irradiation of all proton resonances with a high-power amplitude-modulated and monochromatic

1 H frequency. The irradiation saturates the Boltzman energy levels of the proton spectrum, removing the macroscopic magnetization of all protons and making the scalar proton couplings disappear from the ¹³ C spectrum. There are two main proton decoupling methods: broad band decoupling (BB) and composite pulse decoupling (WALTZ). They differ in the type of modulation of the proton frequencies. Whereas BB uses a continuous irradiation of proton frequencies with a train of rectangular pulses of identical duration and opposite phase (16 ), composite pulse decoupling uses blocks of ¹ H pulses of different pulse lengths and phases (18 ). WALTZ sequences are more efficient than BB, allowing decoupling ¹³ C spectra using less decoupler power and diminishing also the potential dielectric heating of the sample. In addition to the removal of scalar ¹ H couplings, ¹ H irradiation can cause an increase in the ¹³ C signal intensity because of the nuclear overhauser enhancement effect (NOE). The NOE effect is a magnetization transfer process from ¹ H to ¹³ C, which may increase the Figure 1

C MRS spectra (90.55 MHz, 22

C, pH 7.2) of brain extracts after infusion of

(1,2-

C

) glucose (A), (1,2-

C

) acetate (B), or unlabeled glucose (C). Gln C3 (26.87,

J

⫽

1

J

⫽ 34.5), Glu C3 (27.80,

J

⫽

J

⫽ 34.5), Gln C4 (31.60,

J

⫽ 34.5,

J

⫽ 49.0), Glu C4 (34.16,

J

⫽ 34.5,

J

⫽ 51.0), GABA C2 (35.07,

J

⫽ 51.0), Gln C2 (55.0,

J

⫽ 54.96,

1

J

⫽ 34.5), Glu C2 (55.35,

J

⫽ 53.0,

J

⫽ 34.5). Numbers in parenthesis give the chemical shift in ppm, followed by the geminal coupling constant(s) in Hz. Insets in C: Natural abundance contribution of unlabeled GABA in the spectra of vigabatrin treated animals. Ala: alanine, Asp:

aspartate, GABA: ␥ -aminobutyric acid, Gln: glutamine, Glu: glutamate, NAA: N-acetylaspartate,

Tau: taurine. Reproduced from (17) with permission of the publisher.

intensity of a ¹³ C resonance up to threefold in favorable cases. More information on decoupling sequences and NOE enhancement in ¹³ C MRS is

given in the classical monograph by Breitmeier and Voelter (16 ).

Different decoupling schemes are possible that produce proton-coupled or proton-decoupled ¹³ C MRS spectra containing (or not containing) NOE enhancement. If the ¹ H decoupler is not used in con- junction with ¹³ C excitation (see Fig. 3, top panel), a proton-coupled ¹³ C spectrum is obtained. The multi- plet structure of the proton-coupled ¹³ C resonance depends of the number of protons coupled to the observed carbon and follows the rule, 2nI

1. For n

1, a doublet is obtained. If the decoupler is gated only during the acquisition of the ¹³ C FID (inverse gated decoupling), a proton-decoupled spectrum de- void of NOE is obtained. On the contrary, if the decoupler is gated only during the relaxation delay and not during the acquisition (direct gated decou- pling), a proton-coupled NOE-enhanced ¹³ C spectrum is obtained. Inverse-gated decoupling avoids the ef- fects of NOE, whereas direct-gated decoupling gen- erates ¹ H-coupled ¹³ C spectra with complete NOE enhancement. Decoupling during the complete cycle time generates a proton-decoupled NOE-enhanced

13 C spectrum (19).

A different strategy was proposed in 1985 by Roth- man et al. (20). These authors showed that ¹³ C-la- beled metabolites could be detected in vivo with ¹ H sensitivity using a ¹³ C decoupling sequence in con- Figure 2 Homonuclear coupling patterns in two contigu-

ous carbons of the same metabolite molecule. (A)

C-

C;

(B)

C-

C; (C)

C-

C.

Figure 3 Acquisition of

C MRS spectra under different

H-decoupling conditions. Top: proton- coupled

C MRS spectrum; middle: proton-decoupled

C MRS spectrum without NOE; bottom:

proton-decoupled

C MRS spectrum with NOE. The

C or

H symbols at the left of every

sequence indicate RF transmitter or decoupler pulses, respectively.

junction with ¹ H detection (POCE; proton observed carbon edited). Since then, this approach has been used successfully in vivo and in vitro, and several examples are available in the literature (21, 22).

13 C MRS Quantitation

The interpretation of ¹³ C MRS spectra in terms of flux through metabolic pathways requires the quantitation of the ¹³ C incorporated in specific carbons. This is normally done by expressing ¹³ C incorporation as a fractional ¹³ C enrichment in carbon C _i (YC

). The fractional ¹³ C enrichment YC

is defined as the amount of ¹³ C relative to the total carbon ( ¹³ C

¹² C) present in C _i .

YC

⫽

13 C concentration in C

共

¹³ C ⫹ ¹² C

concentration in C

[1]

The measurement of YC

involves (i) the determina- tion of ¹³ C concentration by ¹³ C MRS methods and (ii) the measurement of the total carbon concentration by more conventional techniques (automatic ion ex- change chromatography for amino acids, HPLC, en- zymatic end point methods for other metabolites, and so on).

Special precautions must be taken in the determi- nation of the concentration of ¹³ C in C _i by ¹³ C MRS because ¹³ C signal intensities depend on various fac- tors, in addition to the ¹³ C concentration. Thus, ¹³ C resonances from carbons with the same ¹³ C concen- tration may have different intensities or areas depend- ing on the particular pulsing condition used, the re- laxation behavior of the observed carbon, its NOE effect, and acquisition and repetition times.

The intensity, area, or magnetization M(t) of a ¹³ C carbon resonance is given by Eq. [2] (23):

M

t

⫽ M

0

sin

e

⫺ 1

e

⫺ 1

⫹

⫹ 1

1 ⫺ e

e

[2]

where M(t) is the area of the resonance under the pulsing conditions used, M(0) is the area of the reso- nance under equilibrium magnetization conditions,

is the flip angle used in the ¹³ C pulse,

is the NOE enhancement factor, t is the total cycle time, T ₁ is the longitudinal relaxation time, and a is the acquisition time. M(0) is proportional to the concentration of ¹³ C.

Thus, if the ¹³ C spectrum is acquired under fully relaxed conditions, a comparison of the area M(0) in the sample with the area of the same resonance in a standard solution of known concentration would give

a value for the ¹³ C concentration in the sample. In practice, this procedure is not recommended because T ₁ values of ¹³ C in extracts vary from approximately 3 s in methyl and methylene resonances to 30 s or more in quaternary carbons like those of carboxylic acids. Thus to recover completely the equilibrium magnetization M(0) in the sample, the slowest car- bons would require a total cycle time of approxi- mately 150 s. This is incompatible with reasonable acquisition times. Thus, a compromise is normally adopted and M(0) in the sample and standard solu- tions can be calculated from M(t) measured under partially saturating conditions (3– 6 s total cycle time) and predetermined values of T ₁ and

using the equation described above. To favor quantitation, the influence of the NOE factor is avoided using gated decoupling sequences.

Other ¹³ C MRS quantitation strategies may be used to determine fractional ¹³ C enrichments: (i) direct comparison of the areas of the ¹³ C resonances in the sample spectrum with those of the same resonances in the spectrum of a standard solution of known concen- tration acquired under identical pulsing conditions;

(ii) comparison of the intensities of ¹³ C resonances in the sample spectrum with the intensity of the natural abundance signals (1.1% YC

) from an experiment performed using an unlabeled substrate; and (iii) use of ¹ H MRS to measure the relative area of the ¹³ C satellites to the total area of the ¹ H resonance (see below). Once the ¹³ C concentration is known, the fractional ¹³ C enrichment can be calculated and used in mathematical models of metabolism to determine metabolic flux.

DETERMINATION OF METABOLIC FLUX BY ¹³ C MRS

Kinetic and Steady-State Methods

The main goal of ¹³ C MRS studies is to explore qualitatively or quantitatively metabolic flux through a specific step in a pathway, through a whole pathway, or through a combination of several pathways. To measure metabolic flux by in vivo ¹³ C MRS, a ¹³ C- enriched precursor is administered (infused, peri- fused, and injected) to the biological system. After its administration, ¹³ C atoms from the substrate substi- tute progressively, because of metabolism, the ¹² C atoms previously present in the system. This substi- tution may be monitored kinetically in some favorable cases by ¹³ C MRS techniques (24 –27 ).

Once the kinetics of ¹³ C enrichment are known it

becomes possible to fit the time course of ¹³ C enrich-

ment(s) to a metabolic model containing the fractional

13 C enrichment(s) as the dependent variable(s), time as the independent variable, and the unknown fluxes as the parameters to be fitted. This approach allowed calculating the flux through the tricarboxylic acid cycle and the glutamine cycle in the brain of animals and man (9, 25, 26).

Eventually, an isotopic steady state is achieved after continuous administration of a ¹³ C-enriched sub- strate. Under steady-state conditions, the rate of in- corporation of ¹³ C equals the rate of disappearance, and principles of conservation of mass and isotope are fulfilled (Fig. 4).

These principles state that the total amount of carbon ( ¹³ C

¹² C) and the amount of ¹³ C entering a carbon pool are equal to the total amount of carbon ( ¹³ C

¹² C) and ¹³ C leaving the pool, respectively. It is possible then to set up a series of simultaneous equations relating the fractional ¹³ C enrichments de- termined in every carbon to input output fluxes (see Fig. 4). The unknown fluxes can be found, provided the number of equations is equal to the number of unknowns. This is many times the limiting condition as the number of fractional ¹³ C enrichments deter- mined by ¹³ C MRS or directly measurable fluxes by other conventional methods is small in comparison with the complexity of metabolic transformations in- vestigated. In these cases, the use of simplified models of metabolism has proven to be useful. Examples of the steady-state approach based of isotopic dilution

can be found in the classic work of Chance et al. (27) or studies of neural cell metabolism (28, 29).

The ¹³ C Isotopomer Approach

The methods based exclusively on the determination of fractional ¹³ C enrichment neglect the information provided by spin coupling patterns (see Fig. 2), re- sulting in a significant loss of information. Thus, for a five-carbon metabolite such as glutamate, ¹³ C enrich- ment determinations would produce at most five dif- ferent enrichment values, one for every carbon. These number enrichments would support at most five input- output equations as described in Fig. 4.

The gain in information introduced by the use of spin coupling patterns or isotopic shifts is more easily perceived with the following example. For a molecule such as glutamate containing five carbons, five hydro- gens, four oxygens, and one nitrogen atom, the pos- sibilities of isotopic substitution with ¹³ C, ² H, ¹⁷ O, and ¹⁵ N isotopes are large (Table 1). Glutamate has 32 possibilities of ¹³ C labeling. Each of these possibili- ties is an isotopic isomer (or isotopomer) of glutamate with 32 possibilities of deuteration, four possibilities of ¹⁷ O (or ¹⁸ O) labeling, and two possibilities of nitrogen substitution. Thus, a total of 32,768 isoto- pomers of ( ¹³ C, ² H, ¹⁷ O, ¹⁵ N) glutamate are available for labeling in multinuclear studies. Fortunately, not all of these isotopomers are directly detectable by conventional ¹³ C MRS methods. Only about 40 iso- topomers of ( ² H, ¹³ C) glutamate are easily detected in 1D ¹³ C MRS with combined ¹³ C and ² H labeling;

more would be detectable using more sophisticated 2D or 3D multinuclear MRS techniques. Input-output equations for each of these isotopomers can be written and solved, allowing a significantly larger number of fluxes to be determined than with the conventional fractional enrichment method. However, the large number of equations demands the use of computer programs to calculate and fit simulated ¹³ C spectra to Figure 4 Principles of mass conservation and isotope con-

servation at steady state. YCn, YCm, and YCp refer to fractional

C enrichments in carbons Cn, Cm, and Cp. Fz or Fx and Fy refer to metabolic fluxes leaving and entering the Cp pool, respectively. Mass conservation principle:

Fz ⫽ Fx ⫹ Fy. Isotope conservation principle: YCp.Fz ⫽ YCn.Fx ⫹ YCm.Fy. If Fz is known, Fx and Fy can be calculated if YCp, YCm, and YCn are measured by

C MRS.

Table 1 Possibilities of Isotopic Substitution in Glutamate

Atom Number

Isotopic Substitution

Possibilities of Isotopic Labeling (Isotopomers)

Carbon 5

C 32

Hydrogen 5

C,

H 1,024

Oxygen 4

C,

H,

O(

O) 16,384 Nitrogen 1

C,

H,

O(

O),

15

N

32,768

experimental ones. At least two algorithms are avail- able that perform this task, TCASIM (30) and META- SIM (31). The isotopomer approach was initially de- veloped for the estimation of the contribution of anaplerotic fluxes to the tricarboxylic acid cycle of the heart (32) and has been extended to the study of metabolic compartmentation in heart (33–36) and brain (13, 28, 37– 41). In addition, ² H- ¹³ C couplings have been used in the study of solvent exchange reactions in the perfused liver (42, 43), and ¹⁵ N- ¹³ C and ¹ H- ¹⁵ N couplings in the analysis of ammonia detoxification pathways (44, 45).

APPLICATIONS

Cerebral Metabolic Compartmentation as Revealed by the Use of Substrates

Multiply Enriched in ¹³ C and High- Resolution ¹³ C MRS

A significant number of studies of cerebral metabolic compartmentation have used high-resolution ¹³ C MRS analysis of brain extracts and glucose or acetate contiguously labeled with ¹³ C as cerebral substrates (13, 17, 31, 37– 41). An important basis for the use of this approach is the impressive amount of metabolic information contained in homonuclear spin coupling patterns.

Figure 5 provides an adequate frame to discuss these aspects by showing schematically the cerebral metabolism of (1,2- ¹³ C ₂ ) glucose or (1,2- ¹³ C ₂ ) acetate and their effects on ¹³ C MRS spectra of brain extracts.

(1,2- ¹³ C ₂ ) glucose is transported from plasma to brain cells through glucose transporters and degraded di- rectly to an equimolar mixture of (2,3- ¹³ C ₂ ) and un- labeled pyruvate through the Embden-Meyerhoff pathway. Cytosolic (2,3- ¹³ C ₂ ) pyruvate can be re- duced to (2,3- ¹³ C ₂ ) lactate, transaminated to (2,3-

13 C ₂ ) alanine, or transported to the mitochondria for oxidative metabolism (46). In the mitochondrial ma- trix, (2,3- ¹³ C ₂ ) pyruvate can enter the tricarboxylic acid cycle through pyruvate dehydrogenase as (1,2-

13 C ₂ ) acetyl-CoA or through pyruvate carboxylase as (2,3- ¹³ C ₂ ) oxalacetate. In the first turn, (1,2- ¹³ C ₂ ) acetyl-CoA produces (4,5- ¹³ C ₂ )

ketoglutarate, while (2,3- ¹³ C ₂ ) oxalacetate forms (1,2- ¹³ C ₂ )

ke- toglutarate. The exchange between

-ketoglutarate and glutamate, through aspartate aminotransferase, has been traditionally assumed to be fast compared with the tricarboxyxlic acid flux, and thus glutamate labeling is thought to reflect accurately the labeling in the

-ketoglutarate precursor (22). Thus (4,5- ¹³ C ₂ ) and (1,2- ¹³ C ₂ ) glutamate reflect ¹³ C labeling of the

corresponding

-ketoglutarate precursors. (4,5- ¹³ C ₂ ) and (1,2- ¹³ C ₂ ) glutamate may produce (4,5- ¹³ C ₂ ) and (1,2- ¹³ C ₂ ) glutamine through glutamine synthase or (1,2- ¹³ C ₂ ) and (4- ¹⁴ C) GABA through glutamate de- carboxylase, respectively. If (4,5- ¹³ C ₂ ) or (1,2- ¹³ C ₂ )

␣

-ketoglutarate continue metabolism through the cy- cle, equimolar mixtures of (3,4- ¹³ C ₂ ) and (1,2- ¹³ C ₂ ) or (1- ¹³ C) and (4- ¹³ C) succinates, fumarates, and malates are produced, entering eventually a new turn of the cycle. Several studies of cerebral metabolism used (1,2- ¹³ C ₂ ) acetate as substrate (37– 40). (1,2- ¹³ C ₂ ) acetate is believed to be transported and activated to (1,2- ¹³ C ₂ ) acetyl-CoA only in glial cells, bypassing the pyruvate dehydrogenase and pyruvate carboxylase steps of entry into the tricarboxylic acid cycle.

These ¹³ C labeling patterns are adequately re- flected in the high-resolution ¹³ C MRS spectra of brain extracts by the intensities and multiplicities of the ¹³ C resonances from the individual carbons of glutamate, glutamine, and GABA. This is illustrated in Fig. 1, which shows representative ¹³ C MRS spec- tra obtained from brain extracts after metabolism of (1,2- ¹³ C ₂ ) glucose (A), of (1,2- ¹³ C ₂ ) acetate (B), or of unlabeled glucose (bottom). At steady state, glutamate carbons C4 (34.2 ppm) and C5 (182.2 ppm) reflect labeling of acetyl CoA carbons C2 and C1, whereas glutamate carbons C1 (175.5 ppm), C2 (55.5 ppm), and C3 (27.7 ppm) reflect labeling in oxalacetate carbons C4, C3, and C2, respectively. The relative proportions of (2,3- ¹³ C ₂ ) pyruvate entering the tricar- boxylic acid cycle through pyruvate dehydrogenase or pyruvate carboxylase can be estimated by the relative intensities of the doublet resonances observed in the glutamate C4 and C2 carbons, respectively (47, 48 ).

In addition, ¹³ C MRS spectra of brain extracts provide an excellent tool to study cerebral metabolic compartmentation. This is more easily accomplished through the analysis of the homonuclear spin coupling patterns. Indeed, most of the observed resonances in Fig. 1 display an apparent triplet structure derived from the superposition of doublets and singlets. Dou- blets arise from contiguously ¹³ C-labeled isotopomers of glutamate, glutamine, and GABA, whereas singlets are derived from the corresponding isotopomers with

13 C carbons having only ¹² C neighbors. The presence

of cerebral compartmentation is easily demonstrated

through the analysis of the glutamate and glutamine

C3 and C4 carbon resonances. When (1,2- ¹³ C ₂ ) glu-

cose is the substrate, the different singlet/doublet ratio

in the glutamate and glutamine C3 carbons is incon-

sistent with a single pool of acetyl-CoA and oxalac-

etate condensing at the citrate synthase step of a

unique cerebral tricarboxylic acid cycle. Similarly,

when (1,2- ¹³ C ₂ ) acetate is the substrate, the singlet/

doublet ratio in glutamine C4 is clearly different from that of its precursor glutamate C4, a result inconsistent with the presence of a unique glutamate pool. This difference led to the description of the cerebral pyruvate recycling system, a pathway transforming (1,2- ¹³ C ₂ ) acetyl-CoA units in (1- ¹³ C) and (2- ¹³ C) acetyl-CoA units, which occurs mainly in the synaptic terminals (37, 49). More quantitative interpretations can also be per-

formed with the help of programs for computer-assisted interpretations of ¹³ C MRS spectra (13, 31).

Hydrogen Turnover as Studied by ¹³ C MRS

As described previously, several ¹³ C MRS techniques have been implemented to study the turnover of indi- Figure 5 Cerebral metabolism of (1,2-

C

) glucose and (1,2-

C

) acetate. Filled circles indicate

C.

Empty circles indicate

C. Only the first turn of the tricarboxylic acid cycle is considered. Contiguously labeled isotopomers of glutamate, glutamine and GABA originate the doublet resonances observed in Fig. 1. Single-labeled and natural abundance isotopomers originate singlet resonances in Fig. 1. PDH:

pyruvate dehydrogenase, PC: Pyruvate carboxylase, ME: malic enzyme, PEPCK: phosphoenolpyruvate

carboxykinase, PK: pyruvate kinase. Reproduced from (13) with permission of the publisher.

vidual metabolite carbons by using precursors selec- tively enriched in ¹³ C. Because most metabolite car- bons are attached to one or more hydrogen atoms, ¹³ C MRS could also serve as a valuable tool to analyze hydrogen turnover. This is a much faster process than

13 C turnover and therefore allows studying faster re- actions. Our laboratory has proposed a double-label- ing strategy to investigate hydrogen turnover. Our approach follows by ¹³ C MRS the exchange of pre- existing ¹ H by ² H at the steady state, when metabo- lism occurs in media containing a ¹³ C-labeled sub- strate and ² H ₂ O (40, 42, 43, 50 –52). This is possible because high-resolution ¹³ C MRS is well suited to detect complex deuteration patterns in ¹³ C-labeled metabolites.

The presence of one or more deuterons geminally or vicinally bound to the observed ¹³ C implies the appearance of characteristic ² H isotopic shifts and

2 H- ¹³ C couplings (50, 51, 53). Therefore, coupling patterns to ² H may appear as splitted and/or shifted resonances in relation to the preprotonated ¹³ C reso-

nance. Figure 6 depicts this behavior. If one of the protons directly bonded to ¹³ C is substituted by ² H, the original resonance is splitted into a 1:1:1 triplet (19.21

¹ J _C–H

22 Hz), inducing a geminal upfield isotopic shift (

0.25

1

0.33 ppm). Two or three deuterons bound to the same ¹³ C atom would result in additive isotopic shifts and ² H- ¹³ C couplings patterns of five or seven line multiplets, shifted by

⫺

0.5 or

0.75 ppm, respectively. Even a vicinal deuterium substitution induces smaller and additive upfield isotopic shifts (

0.03

2

0.11 ppm).

Vicinal couplings to deuterium are too small to be resolved, and resonances shifted vicinally maintain the multiplet structure of their geminal couplings.

Thus high-resolution ¹³ C MRS allows determining the number of deuterium replacements, their relative contributions, and their geminal or vicinal location with respect to the observed ¹³ C carbon of a specific

13 C isotopomers through the analysis of shifted and unshifted ( ¹ H, ² H) ¹³ C multiplets.

This kind of approach makes it possible to develop novel procedures to investigate and resolve in time fast metabolic processes, as long as they involve ex- change of hydrogens by deuterons. In our laboratory, we have monitored the dynamics of exchange from specific hydrogens of hepatic glutamate and aspartate with deuterons from intracellular heavy water. With this approach it became possible to observe directly

␣

-ketoglutarate/glutamate exchange and oxalacetate/

aspartate exchange and subcellular compartmentation, by following in a time-resolved manner the sequence of events involved in the traffic of these metabolites through mitochondria and cytosol (42). In this work, the determination of relative ( ¹ H, ² H, ¹³ C) isotopomer populations is conveniently accomplished using a simulation program, which allows the quantitative determination of the relative contributions of the in- dividual ( ¹ H, ² H) ¹³ C multiplets. Figure 7 shows a rep- resentative deconvolution of the C2 carbon resonance of (2- ¹³ C) glutamate. Deconvolution of the C2 carbon resonance revealed contributions from five different ( ¹ H, ² H, ¹³ C) glutamate isotopomers. Notably, these results reveal intracellular glutamate compartmenta- tion and slow

-ketoglutarate/glutamate exchange, a circumstance not considered previously.

Figure 8 illustrates the kinetics of deuteration of the H2 and H3 hydrogens from (2- ¹³ C) glutamate during perfusions with (3- ¹³ C) alanine in Krebs- Ringer bicarbonate (KRB) buffer containing 50%

2 H ₂ O. The faster timescale of hydrogen exchange allows resolving in time cytosolic deuteration, as those of (2- ¹³ C, 3- ² H) glutamate, and mitochondrial deuterations, as that of (2- ¹³ C, 3,3

- ² H ₂ ) glutamate (42).

Figure 6

H-

C multiplets and deuterium-induced isoto- pic shifts in proton-decoupled

C MRS spectra. (A) prep- rotonated carbon, (B) vicinal deuteration, (C–E) geminal deuteration by one, two, or three deuterons. Simulations were performed with the program WINDAISY using the following parameters:

J

⫽ 20.06 Hz; geminal isotopic shift, ⌬

⫽ ⫺ 22.89 Hz/deuteron; vicinal isotopic shift, ⌬

⫽

⫺ 9.89 Hz/deuteron. Reproduced from (50) with permission

of the publisher.

In Vivo ¹³ C MRS and Clinical ¹³ C MRS The possibility to obtain ¹³ C MRS spectra of metab- olism as it proceeds in its own intracellular environ- ment, in animal models or in human beings, repre- sents one of the most attractive possibilities of the ¹³ C

MRS method. Although ¹³ C MRS can be performed in natural abundance conditions, studies with admin- istration of exogenous ¹³ C-enriched precursors are more common. If a ¹³ C label is delivered adequately to an in vivo system, the incorporation of the ¹³ C label Figure 7 Representative deconvolution of the C2 carbon resonances from (2-

C) glutamate into the

contributions of individual (

H,

H)

C multiplets.

C MRS spectra (150.90 MHz, 22°C, pH 7.2) were obtained from the extracts of single livers perfused with 6 mM (3-

C) alanine in 50%

H

O for 15 min (C2 glutamate). Relative contributions of specific isotopomers are given as the fractional contribution of the corresponding multiplet to the total area of the analyzed resonance taken arbitrarily as one.

Simulations were performed with the WINDAISY program. s: singlet, ss: shifted singlets, dss: doubly

shifted singlet, t: triplet, st: shifted triplet. Reproduced from (42) with permission of the publisher.

into some relevant metabolites can be dynamically detected in situ by direct ¹³ C MRS or inverse ( ¹ H) ¹³ C MRS.

A large number of studies using in vivo ¹³ C MRS has been performed in the intact rodent or human brain. Surface coils constitute the most widespread localization technique, improving the sensitivity and avoiding external interferences (54, 55). ¹³ C MRS 3D localization of a selected volume constituted an im- portant methodological challenge approached by sev- eral laboratories during the past decade. Parallel to the development of 3D localization, recent improvements in localized shimming provided an additional contri- bution to the advance of the technique (56, 57 ). Both improvements were fundamental to obtain relevant results, opening the field of in vivo ¹³ C MRS as one of the most powerful techniques in noninvasive stud- ies of neurological processes. Contributions from this technology include (i) observation of the natural abundance signals from some brain metabolites in vivo (e.g., myo-inositol) (54 ); (ii) in situ detection of (1- ¹³ C) glycogen in the intact brain (58 ); and (iii) development of noninvasive studies of cerebral met- abolic compartmentation, as ¹³ C labeling of cerebral glutamate and glutamine can be visualized in vivo (57, 59).

More recently, ¹³ C MRS approaches have contrib- uted significantly to our understanding of neuronal- glial interactions in brain (26, 41, 60, 61). A brief Figure 8 Kinetics of deuteration of the H2 and H3 hydro-

gens from (2-

C) glutamate. Fractional contributions of individual isotopomers in each time point were determined as described in the legend to Fig. 7 and are indicated by specific symbols (insets). Lines represent the best fit to a minimal model of hydrogen exchange. Multiplicities are abbreviated as indicated in Fig. 7. Reproduced from (42) with permission of the publisher.

Figure 9 Metabolic coupling between neurons and glia. Glutamatergic neurotransmission releases glutamate from the synaptic vesicles to the synaptic cleft. Synaptic glutamate may be then recaptured by specific astrocyte transporters (together with 3Na

) to produce glutamine through glutamine synthase. Glial glutamine is released to the extracellular space and taken up by neurons, closing the glutamate-glutamine cycle. ATP required for glutamine synthesis (and Na

extrusion) is produced in the astrocyte by degrading plasma glucose, both oxidatively and nonoxidatively.

Thus, glutamatergic neurotransmission is coupled to the intercellular glutamine cycle and astrocytic

glucose uptake. GLC: glucose, GLN: glutamine, GLU: glutamate, LAC: lactate, TCAg: glial

tricarboxylic acid cycle, TCAn: neuronal tricarboxylic acid cycle.

summary is outlined in Fig. 9. Following activation, neurons release glutamate to the synaptic cleft, which either binds to postsynaptic receptors or is recaptured by surrounding astrocytes in the neuropil. Astrocytic glutamate is transformed into glutamine by the ATP- dependent glutamine synthase, glutamine being ex-

truded from the astrocyte and recaptured by the neu- ron. Together, these processes constitute the glutamate-glutamine cycle that underlies the coupling of neuronal and glial metabolisms. The neuronal and glial tricarboxylic acid cycles and the glutamine cycle and been investigated repeatedly by in vivo and in Table 2 Tricarboxylic Acid Cycle and Glutamate/Glutamine Exchange between the Neuronal Glial

Compartments of the Adult Brain as Calculated with Different Models and Methodologies

Process/Model Garfinkel

Van den Berg and Garfinkel

Kunnecke et al.,

Preece and

Cerdan

Mason et al.,

Sibson et al.,

Shen et al.,

Lebon et

al.

Gruetter et al.

Bluml et al.

Total cerebral TCA cycle flux

( ␮ mol min

g

) 1.05 1.5 1.4

1.6 or 0.7,

0.6,

1.0–0.2,

0.8

0.6 0.84

Neuronal TCA cycle flux

( ␮ mol min

g

) 0.40 1.2 1.0

1.6,

0.6,

1.0–0.2,

0.8

0.6 0.70 Glial TCA cycle flux

( ␮ mol min

g

) 0.65 0.3 0.4 0.14

0.14

Size ( ␮ mol g

)/turnover (min

) of large glutamate

pool 8.8/21.7 7.0/5.7 5.8/5.8 n.a. n.a.

Size ( ␮ mol g

) turnover (min

) of small glutamate

pool 1.7/2.6 1.25/4.16 0.5/1.25 n.d./7.7

n.a.

Net transfer of neuronal gluta- mate to glial compartment

( ␮ mol min

g

) 0.08

0.14 0.1

0.21,

0.40–

0.0,

0.32,

0.3

0.3 (0.2)

n.a.

Net transfer of glial glutamine to neuronal compartment

( ␮ mol min

g

) 0.45 n.a. 0.1

0.21,

0.40–

0.0,

0.32,

0.3

0.3 (0.2)

n.a.

a

Calculated from specific radioactivity measurements in brain extracts obtained after intracraneal injections of various radioactive precursors including (U-

C) glutamate, (U-

C) aspartate,

CO

H

, and

NH

acetate (69).

b

Calculated from specific radioactivity measurements in glutamate, glutamine, and aspartate from mouse brain extracts prepared after intraperitoneal injections of

C-labeled glucose and acetate (70).

c

Relative flux values were calculated as described in (31) and from the relative

C isotopomer populations in glutamate, glutamine, and GABA measured by high-resolution

C NMR in rat brain extracts after infusion of (1,2-

C

) acetate (38).

d

Absolute flux values were determined from the relative values described in note (c) by measuring the absolute rate of GABA accumulation induced by vigabatrin, a selective inhibitor of GABA transaminase (17).

e

Determined in vivo from the kinetics of

C enrichment in glutamate and glutamine C4 carbons from rat or human brain during infusion of (1-

C) glucose, respectively (21, 25).

f

Determined in vivo from the kinetics of

C labeling in glutamate and glutamine C4 (55 ).

g

Determined in vivo from the kinetics of

C enrichment in glutamate and glutamine C4, under different conditions of morphine,

␣ -chloralose, and pentobarbital anesthesia (71).

h

Determined in the human brain in vivo during (1-

C) glucose infusion. Glutamate/glutamine exchange concluded to be stoichiometric 1:1 with CMRglc (72).

i

Determined in the human brain during metabolism of (2-

C) acetate (73).

j

Determined in vivo as the inverse of the rate constant of

C labeling in cerebral glutamate C4 during (1-

C) glucose infusions (74).

k

Determined in the human brain during (1-

C) glucose infusion with a different model than that used in (f–i) (9).

l

Only a fraction of glutamine synthesis considered to be derived from neurotransmitter glutamate. Glutamate/glutamine exchange concluded nonstoichiometric with CMRglc, approaching 0.4 rather than the value of 1 concluded in (f–i) (75 ).

m

Determined in the human brain during (1-

C) acetate metabolism (76).

n

Originally proposed as an ␣ -ketoglutarate exchange between the large and small compartments (69).

n.a.: not applicable, n.d.: not detectable.

Adapted from (13). Reproduced with permission of the publisher.

vitro ¹³ C NMR in a variety of laboratories (26, 41, 60, 61).

Table 2 summarizes values for the tricarboxylic acid cycle flux, as obtained using different in vivo ¹³ C NMR approaches compared with earlier radioactive isotope techniques and in vitro ¹³ C isotopomer ap- proaches. In general, tricarboxylic acid cycle fluxes calculated for the adult rat brain were in the range 1.4 – 0.5

mol

min

¹

g

¹ , depending on the math- ematical model used for the calculation and the degree of anesthesia. Values found in rodent brain were larger than in humans. Table 2 also shows sizes and turnover rates of the glutamate compartments associ- ated to the “large” and “small” glutamate pools. No direct measurements exist to our knowledge of these pool sizes. The values shown are based on the as- sumption that the small glutamate pool accounts for approximately 10% of total cerebral glutamate during (1- ¹³ C) glucose metabolism. Recent evidences indi- cate that the relative sizes and fractional ¹³ C enrich- ments of the large and small glutamate pools could depend on the substrate used (40).

An area of promising development is the clinical use of ¹³ C MRS. The high cost of the MRS equipment and ¹³ C isotopes needed for human studies have re- stricted the use of human ¹³ C MRS to a small number of laboratories. Despite this, several brain disorders have been investigated using ¹³ C MRS, which has already provided new and important clues to clinical diagnosis. The following physiopathological pro- cesses have been investigated: (i) in vivo rate of NAA synthesis (62); (ii) elucidation of mitochondrial brain disorders (63); (iii) the glutamate-glutamine cycle and glutamate neurotransmission in certain diseases (64 );

and (iv) study of hepatic encephalopathy and impaired consciousness (65). Additional reviews describing these topics in more detail can be found in (66 – 68).

CONCLUSION

The results obtained in the past two decades have shown that even with a simple ¹³ C MRS methodology and using conventional substrates such as (1,2- ¹³ C ₂ ) glucose and (1,2- ¹³ C ₂ ) acetate, it is possible to deter- mine quantitatively metabolic flux through important metabolic pathways in vivo and in vitro. The use of heteronuclear labeling strategies and the increasing power of the associated mathematical modeling may increase significantly the number of pathways amena- ble to quantitative analysis by ¹³ C MRS in the near future. Finally, ¹³ C MRS will still need to overcome some important challenges. Among these are the study of intracellular

-ketoglutarate/glutamate, the

study of regional metabolism in different brain areas, the quantitative study of neuronal-glial interactions and their alteration in different pathologies, the devel- opment microscopic ¹³ C MRS approaches to unravel metabolism in a single cell, and the routine applica- tion of in vivo ¹³ C MRS to clinical studies.

ACKNOWLEDGMENTS

The authors are deeply indebted to Mr. Luı´s Lopes da Fonseca for careful reading of the manuscript and to Mr. Javier Pe´rez for careful drafting of the illustra- tions. This work was made possible by grants SAF 2001-2245, SAF 2004-03197 from the Spanish Min- istry of Education and Science and FISss C03/08, G03/155, and C03/10 from the Institute of Health Carlos III to Sebastia´n Cerda´n. Justesa Imagen S.A.

provided a core grant supporting the LISMAR NMR Facility. Tiago B. Rodrigues was supported by a fel- lowship from Fundac¸a˜o para a Cieˆncia e Tecnologia/

Ministe´rio da Cieˆncia e Ensino Superior–Portugal.

REFERENCES

1. Eakin RT, Morgan LO, Gregg CT, Matwiyoff NA.

1972. Carbon-13 nuclear magnetic resonance spectros- copy of living cells and their metabolism of a specifi- cally labeled

C substrate. FEBS Lett 28(3):259 –264.

2. Cerdan S, Seelig J. 1990. NMR studies of metabolism.

Annu Rev Biophys Biophys Chem 19:43– 67.

3. Kunnecke B. 1995. In carbon-13 NMR spectroscopy of biological systems. In: Beckman N. New York: Aca- demic Press pp 159 –267.

4. Morris P, Bachelard H. 2003. Reflections on the appli- cation of C-13-MRS to research on brain metabolism.

NMR Biomed 16(6 –7):303–312.

5. Szyperski T. 1998.

C-NMR, MS and metabolic flux balancing in biotechnology research. Q Rev Biophys 31(1):41–106.

6. Friebolin H. 1991. Basic one- and two-dimensional NMR spectroscopy. Weinheim: VCH Publishers.

7. Canioni P, Alger JR, Shulman RG. 1983. Natural abun- dance carbon-13 nuclear magnetic resonance spectros- copy of liver and adipose tissue of the living rat. Bio- chemistry 22(21):4974 – 4980.

8. Dobbins RL, Malloy CR. 2003. Measuring in-vivo me- tabolism using nuclear magnetic resonance. Curr Opin Clin Nutr Metab Care 6(5):501–509.

9. Gruetter R. 2002. In vivo

C NMR studies of com- partmentalized cerebral carbohydrate metabolism. Neu- rochem Int 41(2–3):143–154.

10. Shulman RG, Rothman DL. 2001.

C NMR of inter-

mediary metabolism: Implications for systemic physi-

ology. Annu Rev Physiol 63:15– 48.

11. Cox IJ. 1996. Development and applications of in vivo clinical magnetic resonance spectroscopy. Prog Bio- phys Mol Biol 65(1–2):45– 81.

12. Jucker BM, Lee JY, Shulman RG. 1998. In vivo

C NMR measurements of hepatocellular tricarboxylic acid cycle flux. J Biol Chem 273(20):12187–12194.

13. Cruz F, Cerdan S. 1999. Quantitative C-13 NMR stud- ies of metabolic compartmentation in the adult mam- malian brain. NMR Biomed 12(7):451– 462.

14. Sonnewald U, Schousboe A, Qu H, Waagepetersen HS.

2004. Intracellular metabolic compartmentation as- sessed by

C magnetic resonance spectroscopy. Neu- rochem Int 45(2–3):305–310.

15. Zwingmann C, Leibfritz D. 2003. Regulation of glial metabolism studied by C-13-NMR. NMR Biomed 16(6 –7):370 –399.

16. Breitmaier E, Voelter W. 1987. Carbon-13 NMR spec- troscopy. High-resolution methods and applications in organic chemistry and biochemistry. Weinheim: VCH Publishers.

17. Preece NE, Cerdan S. 1996. Metabolic precursors and compartmentation of cerebral GABA in vigabatrin- treated rats. J Neurochem 67(4):1718 –1725.

18. Hu H, Shaka AJ. 1999. Composite pulsed field gradi- ents with refocused chemical shifts and short recovery time. J Magn Reson 136(1):54 – 62.

19. van de Ven FJM. 1995. Multidimensional NMR in liquids. Basic principles and experimental methods.

Weinheim: VCH Publishers.

20. Rothman DL, Behar KL, Hetherington HP, den Hol- lander JA, Bendall MR, Petroff OA, Shulman RG.

1985.

H-Observe/

C-decouple spectroscopic mea- surements of lactate and glutamate in the rat brain in vivo. Proc Natl Acad Sci USA 82(6):1633–1637.

21. Mason GF, Behar KL, Rothman DL, Shulman RG.

1992. NMR determination of intracerebral glucose con- centration and transport kinetics in rat brain. J Cereb Blood Flow Metab 12(3):448 – 455.

22. Burgess SC, Carvalho RA, Merritt ME, Jones JG, Mal- loy CR, Sherry AD. 2001.

C isotopomer analysis of glutamate by J-resolved heteronuclear single quantum coherence spectroscopy. Anal Biochem 289(2):187–

195.

23. Post JF, Baum E, Ezell EL. 1992.

C NMR studies of glucose metabolism in human leukemic CEM-C7 and CEM-C1 cells. Magn Reson Med 23(2):356 –366.

24. Beckmann N, Turkalj I, Seelig J, Keller U. 1991.

C NMR for the assessment of human brain glucose me- tabolism in vivo. Biochemistry 30(26):6362– 6366.

25. Mason GF, Gruetter R, Rothman DL, Behar KL, Shul- man RG, Novotny EJ. 1995. Simultaneous determina- tion of the rates of the TCA cycle, glucose utilization, alpha-ketoglutarate/glutamate exchange, and glutamine synthesis in human brain by NMR. J Cereb Blood Flow Metab 15(1):12–25.

26. Rothman DL, Behar KL, Hyder F, Shulman RG. 2003.

In vivo NMR studies of the glutamate neurotransmitter

flux and neuroenergetics: implications for brain func- tion. Annu Rev Physiol 65:401– 427.

27. Chance EM, Seeholzer SH, Kobayashi K, Williamson JR. 1983. Mathematical analysis of isotope labeling in the citric acid cycle with applications to 13C NMR studies in perfused rat hearts. J Biol Chem 258(22):

13785–13794.

28. Cruz F, Villalba M, Garcia-Espinosa MA, Ballesteros P, Bogonez E, Satrustegui J, Cerdan S. 2001. Intracel- lular compartmentation of pyruvate in primary cultures of cortical neurons as detected by

C NMR spectros- copy with multiple

C labels. J Neurosci Res 66(5):

771–781.

29. Portais JC, Schuster R, Merle M, Canioni P. 1993.

Metabolic flux determination in C6 glioma cells using carbon-13 distribution upon [1-

C]glucose incubation.

Eur J Biochem 217(1):457– 468.

30. Jeffrey FM, Storey CJ, Sherry AD, Malloy CR. 1996.

13

C isotopomer model for estimation of anaplerotic substrate oxidation via acetyl-CoA. Am J Physiol 271(4 Pt 1):E788 –799.

31. Kunnecke B, Cerdan S, Seelig J. 1993. Cerebral me- tabolism of [1,2-

C

]glucose and [U-

C

]3-hydroxy- butyrate in rat brain as detected by

C NMR spectros- copy. NMR Biomed 6(4):264 –277.

32. Malloy CR, Sherry AD, Jeffrey FM. 1990. Analysis of tricarboxylic acid cycle of the heart using

C isotope isomers. Am J Physiol 259(3, Pt 2):H987–995.

33. Sherry AD, Zhao P, Wiethoff AJ, Jeffrey FM, Malloy CR. 1998. Effects of aminooxyacetate on glutamate compartmentation and TCA cycle kinetics in rat hearts.

Am J Physiol 274(2 Pt 2):H591–599.

34. Yu X, White LT, Alpert NM, Lewandowski ED. 1996.

Subcellular metabolite transport and carbon isotope ki- netics in the intramyocardial glutamate pool. Biochem- istry 35(21):6963– 6968.

35. Yu X, White LT, Doumen C, Damico LA, LaNoue KF, Alpert NM, Lewandowski ED. 1995. Kinetic analysis of dynamic

C NMR spectra: Metabolic flux, regula- tion, and compartmentation in hearts. Biophys J 69(5):

2090 –2102.

36. Chatham JC, Forder JR, Glickson JD, Chance EM.

1995. Calculation of absolute metabolic flux and the elucidation of the pathways of glutamate labeling in perfused rat heart by

C NMR spectroscopy and non- linear least squares analysis. J Biol Chem 270(14):

7999 – 8008.

37. Cerdan S, Kunnecke B, Seelig J. 1990. Cerebral me- tabolism of [1,2-

C

]acetate as detected by in vivo and in vitro

C NMR. J Biol Chem 265(22):12916 –12926.

38. Chapa F, Kunnecke B, Calvo R, Escobar del Rey F, Morreale de Escobar G, Cerdan S. 1995. Adult-onset hypothyroidism and the cerebral metabolism of (1,2-

13

C

) acetate as detected by

C nuclear magnetic res- onance. Endocrinology 136(1):296 –305.

39. Pascual JM, Carceller F, Roda JM, Cerdan S. 1998.

Glutamate, glutamine, and GABA as substrates for the

neuronal and glial compartments after focal cerebral

ischemia in rats. Stroke 29(5):1048 –1056; discussion 1056 –1047.

40. Chapa F, Cruz F, Garcia-Martin ML, Garcia-Espinosa MA, Cerdan S. 2000. Metabolism of (1-

C) glucose and (2-

C, 2-

H

) acetate in the neuronal and glial compartments of the adult rat brain as detected by (

C,

2

H) NMR spectroscopy. Neurochem Int 37(2–3):217–

228.

41. Garcia-Espinosa MA, Rodrigues TB, Sierra A, Benito M, Fonseca C, Gray HL, Bartnik BL, Garcia-Martin ML, Ballesteros P, Cerdan S. 2004. Cerebral glucose metabolism and the glutamine cycle as detected by in vivo and in vitro

C NMR spectroscopy. Neurochem Int 45(2–3):297–303.

42. Garcia-Martin ML, Garcia-Espinosa MA, Ballesteros P, Bruix M, Cerdan S. 2002. Hydrogen turnover and subcellular compartmentation of hepatic [2-

C]gluta- mate and [3-

C]aspartate as detected by

C NMR.

J Biol Chem 277(10):7799 –7807.

43. Moldes M, Cerdan S, Erhard P, Seelig J. 1994.

H-

H exchange in the perfused rat liver metabolizing [3-

C]alanine and

H

O as detected by multinuclear NMR spectroscopy. NMR Biomed 7(6):249 –262.

44. Lapidot A, Gopher A. 1997. Quantitation of metabolic compartmentation in hyperammonemic brain by natural abundance

C-NMR detection of

C-

N coupling patterns and isotopic shifts. Eur J Biochem 243(3):597–

604.

45. Preece NE, Cerdan S. 1993. Determining

N to

N ratios in biofluids by single-pulse

H nuclear magnetic resonance. Anal Biochem 215(2):180 –183.

46. Clarke DD, Lajhta AL, Maker HS. 1989. Intermediary metabolism. In: Siegel G, Agranoff B, Albers RW, Molinoff P, editors. Basic neurochemistry, Vol. 1. New York: Raven Press pp 541–564.

47. Lapidot A, Gopher A. 1994. Cerebral metabolic com- partmentation. Estimation of glucose flux via pyruvate carboxylase/pyruvate dehydrogenase by

C NMR iso- topomer analysis of D-[U-

C]glucose metabolites.

J Biol Chem 269(44):27198 –27208.

48. Malloy CR, Sherry AD, Jeffrey FM. 1987. Carbon flux through citric acid cycle pathways in perfused heart by

13

C NMR spectroscopy. FEBS Lett 212(1):58 – 62.

49. Cruz F, Scott SR, Barroso I, Santisteban P, Cerdan S.

1998. Ontogeny and cellular localization of the pyru- vate recycling system in rat brain. J Neurochem 70(6):

2613–2619.

50. Garcı´a-Martı´n ML, Ballesteros P, Cerdan S. 2001. The metabolism of water in cells and tissues as detected by NMR methods. Prog NMR Spectrosc 39:41–77.

51. Hansen PE. 1983. Isotope effects on nuclear shielding.

Ann Rep NMR Spectrosc 15:105–234.

52. Rodrigues TB, Gray HL, Benito M, Garrido S, Sierra A, Geraldes CF, Ballesteros P, Cerdan S. 2005. Futile cycling of lactate through the plasma membrane of C6 glioma cells as detected by (

C,

H) NMR. J Neurosci Res 79(1–2):119 –127.

53. Dziembowska T, Hansen PE, Rozwadowski Z. 2004.

Studies based on deuterium isotope effect on

C chem- ical shifts. Prog Nucl Magn Reson Spectrosc 45:1–29.

54. Gruetter R, Rothman DL, Novotny EJ, Shulman RG.

1992. Localized

C NMR spectroscopy of myo-inosi- tol in the human brain in vivo. Magn Reson Med 25(1):204 –210.

55. Sibson NR, Dhankhar A, Mason GF, Behar KL, Roth- man DL, Shulman RG. 1997. In vivo

C NMR mea- surements of cerebral glutamine synthesis as evidence for glutamate-glutamine cycling. Proc Natl Acad Sci U S A 94(6):2699 –2704.

56. Gruetter R, Boesch C. 1992. Fast, noniterative shim- ming of spatially localized signals—in vivo analysis of the magnetic-field along axes. J Magn Reson 96(2):

323–334.

57. Gruetter R. 1993. Automatic, localized in vivo adjust- ment of all first- and second-order shim coils. Magn Reson Med 29(6):804 – 811.

58. Choi IY, Tkac I, Gruetter R. 2000. Single-shot, three- dimensional “non-echo” localization method for in vivo NMR spectroscopy. Magn Reson Med 44(3):387–394.

59. Gruetter R, Novotny EJ, Boulware SD, Mason GF, Rothman DL, Shulman GI, Prichard JW, Shulman RG.

1994. Localized

C NMR spectroscopy in the human brain of amino acid labeling from D-[1-

C]glucose.

J Neurochem 63(4):1377–1385.

60. Rodrigues TB, Cerdan S. 2005. The cerebral tricarbox- ylic acid cycles. In: Gibson G, Dienel D, editors. Neural energy utilization, Vol. 5. Handbook of neurochemistry and molecular neurobiology. 3rd ed. New York: Klu- wer Academic/Plenum. In press.

61. Sierra A, Rodrigues TB, Benito M, Ballesteros P, Gar- cia-Espinosa MA, Garcia-Martin ML, Cerdan S. 2005.

Pyruvate oxidation and mitochondrial substrate trans- port in the central nervous system. In: Gruetter R, editor. Neural metabolism in vivo, Vol. 4. Handbook of neurochemistry and molecular neurobiology. 3rd ed.

New York: Kluwer Academic/Plenum Publisher. In press.

62. Moreno A, Ross BD, Bluml S. 2001. Direct determi- nation of the N-acetyl-L-aspartate synthesis rate in the human brain by

C MRS and [1-

C]glucose infusion.

J Neurochem 77(1):347–350.

63. Bluml S, Moreno A, Hwang JH, Ross BD. 2001. 1-

C glucose magnetic resonance spectroscopy of pediatric and adult brain disorders. NMR Biomed 14(1):19 –32.

64. Lin AP, Shic F, Enriquez C, Ross BD. 2003. Reduced glutamate neurotransmission in patients with Alzhei- mer’s disease—an in vivo

C magnetic resonance spectroscopy study. Magma 16(1):29 – 42.

65. Bluml S, Moreno-Torres A, Ross BD. 2001.

[1-

C]glucose MRS in chronic hepatic encephalopathy in man. Magn Reson Med 45(6):981–993.

66. Gruetter R, Adriany G, Choi IY, Henry PG, Lei HX, Oz GL. 2003. Localized in vivo C-13 NMR spectroscopy of the brain. NMR Biomed 16(6 –7):313–338.

67. de Graaf RA, Mason GF, Patel AB, Behar KL, Roth-

man DL. 2003. In vivo

H-[

C]-NMR spectroscopy of cerebral metabolism. NMR Biomed 16(6 –7):339 –357.

68. Ross B, Lin A, Harris K, Bhattacharya P, Schweinsburg B. 2003. Clinical experience with C-13 MRS in vivo.

NMR Biomed 16(6 –7):358 –369.

69. Garfinkel D. 1966. A simulation study of the metabo- lism and compartmentation in brain of glutamate, as- partate, the Krebs cycle, and related metabolites. J Biol Chem 241(17):3918 –3929.

70. van den Berg CJ, Garfinkel D. 1971. A stimulation study of brain compartments. Metabolism of glutamate and related substances in mouse brain. Biochem J 123(2):211–218.

71. Sibson NR, Dhankhar A, Mason GF, Rothman DL, Behar KL, Shulman RG. 1998. Stoichiometric cou- pling of brain glucose metabolism and glutamatergic neuronal activity. Proc Natl Acad Sci U S A 95(1):

316 –321.

72. Shen J, Shungu DC, Rothman DL. 1999. In vivo chem- ical shift imaging of gamma-aminobutyric acid in the human brain. Magn Reson Med 41(1):35– 42.

73. Lebon V, Petersen KF, Cline GW, Shen J, Mason GF, Dufour S, Behar KL, Shulman GI, Rothman DL. 2002.

Astroglial contribution to brain energy metabolism in humans revealed by

C nuclear magnetic resonance spectroscopy: Elucidation of the dominant pathway for neurotransmitter glutamate repletion and measurement of astrocytic oxidative metabolism. J Neurosci 22(5):

1523–1531.

74. Mason GF, Rothman DL, Behar KL, Shulman RG.

1992. NMR determination of the TCA cycle rate and alpha-ketoglutarate/glutamate exchange rate in rat brain. J Cereb Blood Flow Metab 12(3):434 – 447.

75. Gruetter R, Seaquist ER, Ugurbil K. 2001. A mathe-

matical model of compartmentalized neurotransmitter metabolism in the human brain. Am J Physiol Endocri- nol Metab 281(1):E100 –112.

76. Bluml S, Moreno-Torres A, Shic F, Nguy CH, Ross BD. 2002. Tricarboxylic acid cycle of glia in the in vivo human brain. NMR Biomed 15(1):1–5.

BIOGRAPHIES

Tiago B. Rodrigues graduated as a bio- chemist from the University of Coimbra, Portugal, in 2000. In 2001, he visited the Mary Nell and Ralph B. Rogers Magnetic Resonance Center at the University of Texas Southwestern Medical Center at Dallas where he investigated the Ca

regulation of the tricarboxylic acid cycle by

C isoto- pomer kinetic analysis. Since 2002, he has been a Ph.D. student at the University of Coimbra and also devel- oping the lab work in the Institute of Biomedical Research “Alberto Sols” CSIC/UAM. His research is focused on lactate turnover in neural cells and transformed cell lines as monitored with double isotope-labeled NMR methodologies.

Sebastia´n Cerda´n is the Director of the In- stitute of Biomedical Research “Alberto Sols” CSIC/UAM as well as the Laboratory for Imaging and Spectroscopy by Magnetic Resonance. He has contributed a variety of

13

C NMR and multiple-isotope approaches

to hepatic and cerebral metabolisms since the

early 1980s. His current main interests in-

volve the study of metabolic compartmenta-

tion in liver or brain and the development of fast hydrogen turnover

measurements to investigate intercellular and subcellular lactate

and glutamate metabolisms in these organs.